基于运动特征增强双流网络的视频行为识别  

Video Action Recognition Based on Sport Feature Enhancement in Two-stream Network

作　　者：曹若琛 冯秀芳[1] 赵晨[1] CAO Ruo-chen;FENG Xiu-fang;ZHAO Chen(School of Software,Taiyuan University of Technology,Jinzhong 030600,China)

机构地区：[1]太原理工大学软件学院,晋中030600

出　　处：《科学技术与工程》2025年第4期1540-1546,共7页Science Technology and Engineering

基　　金：山西省重点研发计划(202102020101007)。

摘　　要：为解决目前行为识别中双流网络对运动特征提取的不充分导致识别准确度低的问题,通过提出基于运动特征增强双流网络的行为识别方法提高准确率。该网络分为空间流和时间流,空间流网络和时间流网络结构相同,输入不同。空间流网络输入为视频帧序列,而时间流网络输入为视频帧差序列。网络结构以Resnet50为骨干网络,将3×3卷积替换为所提出的全局运动特征模块和局部运动特征模块,充分提取视频运动信息,最终将空间流和时间流结合输出结果。结果表明:该模型在UCF101和HMDB51数据集上准确率达到96.8%和75.3%,与传统算法相比有一定优越性。To solve the problem of insufficient extraction of sport features by dual stream networks in current action recognition,which leads to low recognition accuracy,a action recognition method based on sport feature enhancement two-stream networks was proposed to improve accuracy.The network was divided into spatial stream and temporal stream,with the same structure but different inputs.The input of the spatial stream network was a video frame sequence,while the input of the temporal stream network was a video frame difference sequence.The network structure used Resnet50 as the backbone network,replacing the 3×3 convolution with the proposed global sport feature module and local sport feature module,fully extracting video sport information,and finally combining spatial and temporal stream to output the results.The results show that the accuracy of the model on the UCF101 and HMDB51 datasets reaches 96.8%and 75.3%,which is superior to traditional algorithms.

关 键 词：行为识别 深度学习 运动特征 双流网络 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]